# EOSIO API Healthcheck for HAProxy This program implements EOSIO healthcheck for HAProxy over TCP. ## TCP Protocol The protocol is simple and has 4 rules. 1. Data is transmitted in `packets` encoded in ascii and ends with newline `\n`. 2. There are two types of packages: `Request` and `Response`. Each `Request` has exactly one `Response`. 3. Each parameter inside a `Request` is separated by `:` 4. Each response contains exactly one `status code` (see below) ### Request The following parameters are supported in a request and are ordered from first to last below: | # | Name | Required | Description | | - | ---------- | -------- | ---------------------------------------------------------------------- | | 1 | Host | Yes | IP/Hostname to the api. | | 2 | Port | No | Port number to the api (default `80`) | | 3 | num_blocks | No | Number of blocks the api can drift before reported `down` (default 10) | | 4 | hyperion | No | `1` if hyperion healthcheck should be used (default `0`) | ### Response The api can respond with exactly one `status code`. See [HAproxy documentation](https://cbonte.github.io/haproxy-dconv/1.7/configuration.html#5.2-agent-check) for more information | code | Description | | --------- | ---------------------------------------------------------- | | `up` | Api is healthy | | `down` | Api is not healthy | | `failed` | The program failed to read the status from the api. | | `maint` | Api is set in maintenance mode (not used) | | `ready` | Api is ready again after being in `maint` state (not used) | | `stopped` | Api has been stopped manually (not used) | ## Author Henrik Hautakoski - [henrik@eossweden.org](mailto:henrik@eossweden.org)